Our previous clinical and pathologic investigations of erythema multiforme (EM) have demonstrated that herpes simplex virus is closely linked to EM and that herpes-associated EM (HAEM) is the most commonly observed form of EM. We have further obsrved that recurrent EM is virtually always HSV associated. In the skin obtained from active lesions of EM we have detected a HSV specific antigen. The antigen, glycoprotein B (gB) is found within human keratinocytes (HK), which is the site of injury in HAEM. We now propose to extend our clinical and pathologic studies to the molecular level; using molecular virology techniques to investigate the role of HSV in EM. We propose to determine if HSV disseminates in HAEM by examining buffy coats of subjects, before, during and after episodes of EM. We also wish to determine if HSV is latent in the skin of patients with HAEM by examination of EM skin lesions and of previously involved skin sites after the EM has cleared. To complete these studies we have prepared a riboprobe which includes the region of viral DNA that encodes for gB. The probe is HSV-specific and highly sensitive to the picogram level. We will use the roboprobe for in situ hybridization techniques and dot-blot hybridization to identify HSV. We hypothesize that the cutaneous injury is not merely non- specific hypersensitivity as previously thought, but HSV-specific involving the host response to HSV antigens expressed on HK. We proposed to examine the mechanism of injury using several models of cytotoxicity. Finally, we propose to study the effect of therapeutic intervention in EM using an anti-HSV drug, acyclovir, to suppress EM.